Explaining Life by Explaining it Away — February 6th, 2010 by Douglas Axe
Reading Stuart Kauffman’s book At Home in the Universe some fourteen years ago, I encountered the following:
I hope to persuade you that life is a natural property of complex chemical systems, that when the number of different kinds of molecules in a chemical soup passes a certain threshold, a self-sustaining network of reactions—an autocatalytic metabolism—will suddenly appear. Life emerged, I suggest, not simple, but complex and whole, and has remained complex and whole ever since… The secret of life, the wellspring of reproduction, is not to be found in the beauty of Watson-Crick pairing, but in the achievement of catalytic closure. [1]
When chemicals react, they produce different chemicals. So the idea here—call it Kauffman’s conjecture—was that mixtures with a sufficient number of different chemicals are bound to give rise to local compositions that continually replenish themselves through a self-catalyzed network of chemical reactions. Those special compositions would typically differ from the original mixture, but since they make more of themselves, they should be able to ‘grow’ by establishing themselves repeatedly in local pockets. The ability to propagate in this way, if proven, would be something like reproduction, only at the low level of chemical composition rather than at the high level of organismal form.
It was clear enough to me why Kauffman and others liked this idea. If some kind of reproduction and inheritance could conceivably be achieved in systems that are much, much simpler than anything we think of as living, then maybe scientists were making the problem of explaining life much, much harder than it really needed to be.
I saw that, but ultimately I still found the proposal unpersuasive. It’s not that I had any problem with autocatalysis or reaction networks, but rather that equating such things with life seemed like gross oversimplification—like equating Kauffman’s book with ink and paper. Even if there was a grain of truth in his proposal, it left too much unexplained.
At the time, my work was focusing on the profound differences between the simple catalysis caused by small molecules and the elaborately orchestrated and stunningly efficient catalysis achieved by enzymes—the catalysts of life. Kauffman was equating complexity with the sheer numbers of chemical species and reactions, whereas my concern was with the mode of reaction. Since Kauffman’s model employed reactions that were fundamentally simple, with no discernable prospect of rising above this, I saw no satisfactory connection between his model and life.
But the difficulty of explaining life’s origin makes even hints of progress a big deal, and many saw in Kauffman’s simple model the potential for something bigger. The reason for their optimism, I think, was expressed by Daniel Dennett around the time of Kauffman’s writing: “Evolution will occur whenever and wherever three conditions are met: replication, variation (mutation), and differential fitness (competition).” [2] The hope was that if autocatalytic networks can deliver those three things, then whatever they lack in comparison to modern life they can acquire through progressive evolution.
I think it’s fair to say that the optimism has faded as the years have passed without anything like a convincing demonstration—at least nothing that could be called “autocatalytic metabolism.” Now it seems things may be drawing to a close with a new paper by Vasas, Szathmáry, and Santos. [3] Their work calls this whole notion of life starting with raw metabolism into question by seriously undermining the biological relevance both of Kauffman’s conjecture and of Dennett’s dictum.
The paper’s title is a diplomatic statement of its main conclusion: Lack of evolvability in self-sustaining autocatalytic networks constrains metabolism-first scenarios for the origin of life. [4] It becomes clear on reading the paper that the word constrains is here being used euphemistically. After testing the effect of fitness on the evolution of their model compositional assemblies, they report that “some slight relative increases and decreases in their replication-mutation equilibrium frequencies are detected, but the effects are so minor that it is hard to think of any evolutionary relevance.” The problem is that the behavior of the whole system is almost completely determined by the inherent chemistry, leaving no room for selection to do anything interesting.
The paper’s conclusions speak to the whole field of pre-biotic evolution:
There is always a danger in using terms that acquire implicit theoretical content as, for example, the term evolution that in biology is normally used to mean Darwin’s theory of evolution by natural selection… Restricting ourselves to this usage of the word “evolution,” the computed population dynamics of growing noncovalent molecular assemblies that undergo splitting when a critical size is reached clearly illustrates that compositional assemblies do not evolve.
And as a finale:
We now feel compelled to abandon compositional inheritance as a jumping board toward real units of evolution.
Now, I’m not suggesting that they favor intelligent design. I’m simply pointing out that yet another once-favored alternative to ID seems to have been reduced to an epitaph. Moreover, while the excellent work of Vasas, Szathmáry, and Santos brings clarity to the situation, we should have known this day would come all along. How? Because whether or not there was a jumping board to real evolution, it’s virtually certain that there isn’t a jumping board to real life.
Think of it this way. If no conceivable mixture of small molecules provides even a faint hope for the emergence of metabolism catalyzed by genetically encoded enzymes, then whatever these mixtures may or may not do, they can’t explain life as we see it. And as the evidence now stands, one would be hard pressed to argue that there is even a faint hope. Vasas, Szathmáry, and Santos have urged the origin-of-life community to keep the true essence of Darwinian evolution in mind, which is clearly important. Even more important, though, is the need to keep the true essence of life in mind.
[3] doi:10.1073/pnas.0912628107
[4] The actual title of the PNAS early edition has the word ‘constraints’ in place of ‘constrains’, which appears to be a typographical error. PubMed lists the title as: Lack of evolvability in self-sustaining autocatalytic networks: A constraint on the metabolism-first path to the origin of life.